A kubernetes node shows failed task “The object ‘vim,Datastore: ds:///vmfs/volumes/<UUID>’ has already been deleted or has not been completely created.”
Article ID: 423793
Updated On:
Products
Issue/Introduction
A Kubernetes worker node task is failing with a "NotFound" error when attempting to access a datastore volume. The worker node shows a failed task on the vCenter UI with the error: “The object ‘vim,Datastore: ds:///vmfs/volumes/<UUID>’ has already been deleted or has not been completely created.”
Environment
VMware vSphere 8.x
VMware vSphere Container Storage Plug(CNS/CSI)
Cause
This error occurs because the Persistent Volumes (PVCs) were left on the original datastore during the migration (Storage+Compute). Since the Kubernetes node is now running on a host that doesn't have access to that specific datastore, the volume attachment fails.
Resolution
Note: Please consult your Kubernetes vendor documentation or support before migrating Kubernetes nodes.
Migrate the Kubernetes node VM (both compute and storage) back to the ESXi host that has access to the source datastore where the PVC is stored. This co-locates the VM with its required storage on a host with the necessary connectivity to fix the reported issue.
Feedback
